Kubescape results are interrupted by storage pod preemption #5996

🤖 Generated by the Agentic Engineer

Description

Kubescape's results API can become unavailable when a scheduler preemption moves its storage pod and the volume takes time to reattach. This interrupts posture, vulnerability and runtime coverage reads.

Steps to Reproduce

1. In an isolated test environment, schedule the results store with its existing storage and resource settings on a constrained worker.
2. Admit a higher-priority workload and observe the results API through the store's reschedule and volume reattachment.

Expected Behavior

The results API stays within its recovery budget, and volume-backed workloads can still schedule.

Actual Behavior

The results API became unavailable during the observed reschedule and recovered after the volume attached.

Acceptance Criteria

• Reproduce the interruption with isolated data and evaluate placement under measured capacity.
• Preserve the single writer, existing volume, scan coverage, and scheduling of disk-bound workloads.
• Verify the three production coverage reads after a guarded rollout and reschedule.

Part of #2447